The long-range goal of the proposed research is to develop effective, evidence-based interventions for remediating memory deficits associated with psychiatric and neurological disorders. A growing body of research with functional and structural neuroimaging methods demonstrates that the human brain has a remarkable capacity for "plasticity", or experience-dependent change, even in adulthood. Training-related changes in patterns of brain activity have been found to result from behavioral interventions to improve memory performance, but presently very little is known about the factors that promote training gains, and the transfer, or generalization, of those gains to other tasks. This research area holds great promise for impacting public health, through eventual clinical applications for disorders such as schizophrenia. Thus, understanding and better characterizing the neuroplasticity associated with behavioral changes in memory training studies, and identifying brain-based principles for the transfer of training, will be critically important for the translational development of interventions to improve cognitive performance in patient populations. In the context of a large memory training intervention with 8 training sessions and pre- and post-training fMRI scanning sessions, the proposed research will manipulate the variables of training task demands (adaptive vs. non-adaptive difficulty) and targets of training (memory strategies vs. memory processes). 90 healthy young adults will participate, with 30 assigned to each of three training conditions. In each condition, the training sessions involve three working memory (WM) training tasks, one serving as a "criterion task" which is administered during the pre- and post-training sessions (in the scanner), as well during the training sessions. Transfer of training will be assessed with benchmark tasks administered at the pre- and post-training sessions, some performed in the scanner. The effects of adaptive difficulty on memory process training will be examined by directly comparing two conditions that target the same executive function for training, differing only in whether the difficulty level of the training tasks is raised incrementally based on individual performance (adaptive) or the difficulty level is held constant throughout all 8 training sessions (non-adaptive). Additionally, the adaptive difficulty condition that targets memory process training (WM updating) will be compared to an adaptive difficulty condition that targets memory strategy instruction (relational encoding). I will test the following theoretically-driven hypotheses: 1) that training under conditions of adaptive difficulty will be associated with broader transfer and with greater changes in task-related brain activity than training under non-adaptive conditions;2) that interventions designed to target memory processes ("process training") will be associated with broader transfer, and with training-related activation changes in different brain regions, than interventions designed to target memory strategies ("strategy instruction");and 3) that pre-training neural overlap will determine which untrained benchmark tasks demonstrate training-induced behavioral plasticity and neural plasticity.